Low-cost, high-speed identification of counterfeit antimalarial drugson paper

Myra T. Koesdjojo, Yuanyuan Wu, Anukul Boonloed, Elizabeth M. Dunfield,Vincent T. Remcho n

Department of Chemistry, Oregon State University, 153 Gilbert Hall, Corvallis, OR 97331, United States

a r t i c l e i n f o

Article history:Received 27 March 2014Received in revised form21 May 2014Accepted 23 May 2014Available online 27 June 2014

Keywords:Paper microfluidicsArtesunateMalariaFabricationColorimetric assay

a b s t r a c t

With the emergence of artesunate antimalarial counterfeiting in Southeast Asia and sub-Saharan Africa,we present the production of a rapid, inexpensive and simple colorimetric-based testing kit for thedetection of counterfeit artesunate in order to preserve life and prevent the development of multi-drugresistant malaria. The kit works based on paper microfluidics which offer several advantages overconventional microfluidics, and has great potential to generate inexpensive, easy-to-use, rapid anddisposable diagnostic devices. Here, we have developed a colorimetric assay that is specific to artesunateand turns yellow upon addition of the sample. The test can be done within minutes, and allows fora semi-quantitative analysis of the artesunate tablets by comparing the developed yellow color on thepaper test to a color-coded key chart that comes with the kit. A more accurate and precise analysis isdone by utilizing a color analyzer on an iPhone camera that measures the color intensity of thedeveloped color on the paper chip. A digital image of the chip was taken and analyzed by measuring theaverage gray intensity of the color developed on the paper circle. A plot of the artesunate concentrationversus the average gray scale intensity was generated. Results show that the intensity of the yellow colordeveloped on the paper test was consistent and proportional to the amount of artesunate present in thesample. With artesunate concentrations ranging from 0.0 to 20 mg/mL, a linear calibration plot wasobtained with a detection limit of 0.98 mg/mL.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Artesunate, the most prevalent artemisinin derivative, is ananti-malarial drug that is critical in the treatment of severemalaria infections and multi-drug resistant infections caused byPlasmodium falciparum. The mortality rate of untreated severemalaria is reported to be 100%, though the mortality rate drops to15–20% with proper treatment [1]. The rapid therapeutic responseand high tolerability have made these artemisinin drugs essentialin treating the infection, and have been used extensively inSoutheast Asia. Thus, it is essential that those with severe malariabe treated rapidly with this life-saving artesunate drug. Owing tothe emergence of artesunate anti-malarial drug counterfeitingin Southeast Asia and sub-Saharan Africa, we developed a rapid,inexpensive, simple colorimetric-based test kit for the detection ofcounterfeit artesunate. The artesunate detection kit is a novelproduct and well suited for resource-poor settings, and improves

on the current lab-based testing practice with a portable, semi-quantitative, field-suitable test that removes the manual stepsrequired during sample processing. The kit will bring an innova-tive approach to address a global health challenge facing thedeveloping world, where counterfeiting of artesunate has becomea significant threat to our ability to counter the spread of malaria.

In Southeast Asia and sub-Saharan Africa there is a growingproblem of counterfeiting of artesunate drugs. At a cost of appro-ximately US$ 1–2 per adult treatment, the relatively high pricedartesunate drugs have stimulated the growth of counterfeit artesu-nate production. Such counterfeits contain sub-therapeutic quan-tities of artesunate or no active drug at all, and have resulted in lossof life due to inadequate treatment of malarial infection [2,3].According to The World Health Organization, counterfeiting ofanti-malarial drugs leads to an estimated 20% of the 1 milliondeaths a year from malaria [4]. It has been reported that artesunatecounterfeits comprise 38–53% of the drugs in Cambodia, Laos,Myanmar, Thailand, and Vietnam [5]. Artesunate is also used totreat multi-drug resistant P. falciparum in locations where malariahas developed resistance to previously used drugs. While artesu-nate is still largely effective against multi-drug resistant malaria,

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/talanta

Talanta

http://dx.doi.org/10.1016/j.talanta.2014.05.0500039-9140/& 2014 Elsevier B.V. All rights reserved.

n Corresponding author. Tel.: þ1 541 737 8181; fax: þ1 541 737 2062.E-mail address: Vincent.Remcho@oregonstate.edu (V.T. Remcho).

Talanta 130 (2014) 122–127

sub-therapeutic treatment with such counterfeit drugs can pro-mote the emergence of new multi-drug resistant malaria. Thus,counterfeiting artesunate not only leads directly to patient mortalityby not obtaining adequate treatment, but it may exacerbate theproblem further via the development of drug resistant malaria [6].

Identification of authentic artesunate is therefore essential toprevent fatalities. Detection of artesunate by the standard spectro-photometric method is proven to be challenging since it absorbs lightonly at low wavelengths, has a low molar extinction coefficient,and retains no distinct UV/vis spectra or fluorescent properties.Artesunate can be identified and quantified by High PerformanceLiquid Chromatography (HPLC); however, such instrumentation isnot typically available in developing countries [7–9]. More labora-tories are needed worldwide to test the authenticity of the drugs.Only one out of 47 countries in Africa where malaria is endemic isequipped to perform the test.

The US Centers for Disease Control and Prevention (CDC) hasdeveloped a colorimetric test for artesunate that is both rapid andinexpensive in the laboratory. The test involves reaction of analkali decomposition product of artesunate with the diazoniumsalt solution of Fast Red TR [5]. In the presence of artesunate,a yellow color is produced [10,11]. We developed a kit that containsa paper-based test strip that carries out the colorimetric assay uponaddition of the artesunate sample. Hence, the manual multi-stepprocesses of adding and mixing of the chemicals involved in thecurrently used colorimetric artesunate test are removed, and thetest can be rapidly performed in locations such as remote hospitals,pharmacies or in the field where resources are limited.

There have been several recent advancements in the fabricationof paper microfluidic chips. Fabrication of paper-based chips canbe done by photolithography [12], wax printing [13,14], plasmaetching [15], inkjet etching [16], and use of a cutting plotter [17].Unlike traditional microfluidics, which often require pumps tomove fluid through the microfluidic channels, paper microfluidicscan be performed without such instrumentation due to the flow offluid being driven by capillary action through the paper. Hence,paper-microfluidics are well suited for use in point-of-care diag-nostics and developing countries where expensive instrumenta-tion is not available.

The paper test is divided into layers with the appropriate driedreagents, and the easily prepared sample solution will flowthrough the regions via capillary action in order to carry out theartesunate detection test. The chemicals required to perform theassay cost approximately US$0.02 per test, which makes it a verypractical solution to detect counterfeits [18]. Since the reagents forthe colorimetric test are stored on paper in dry form, they aremore stable and easier to transport, which provides advantagesof easy handling and longer shelf life. Furthermore, while mostcounterfeiters are producing fake drugs that lack artesunate, somedrugs are made with significantly lower active ingredients, whichare incapable of killing all the parasites. This is an effort by thecounterfeiters to misleadingly pass the test when a qualitative(positive/negative) evaluation is made to determine the presenceof artesunate in the drug. Our test kit allows for quantitativeanalysis of artesunate tablets by providing a key that comes alongwith the kit and allows for comparison of the developed yellowcolor with the intensity of the yellow color corresponding to theapproximate concentration of artesunate.

Our test device is capable of rapidly and inexpensively detect-ing artesunate in commercially available drugs. In this work, wedeveloped a disposable paper-based microfluidic test equippedwith a color indicator that turns yellow in the presence ofartesunate. The chip is constructed from paper materials, withall the required chemicals and reagents to perform the test storeddirectly in the device. The artesunate sample can then be applieddirectly onto the test strip/device for measurement. The color willfully develop within minutes so that the user can determinewhether artesunate is present in the drug formulation at atherapeutic dose or not. The kit provides a color-coded chart,similar to that accompanying pH paper, which is used to deter-mine the relative concentration of artesunate in the tablet. Thepaper-based assay, together with the accompanying color chartprovides a reliable semi-quantitative measurement of artesunatein a tablet. The rapid, simple, and inexpensive test is especiallyuseful when used as a screening tool for counterfeits in remoteareas.

For a more accurate quantitative measurement, we utilizeda color analyzer on a camera phone to take a digital image of the

Fig. 1. Schematic diagram of the paper test kit.

M.T. Koesdjojo et al. / Talanta 130 (2014) 122–127 123

chip and analyze the color developed on the paper test bymeasuring its gray intensity. To accomplish this, an iPhone cameraequipped with a color analyzer app called ColorAssist is used toanalyze the RGB value of the yellow color produced, and convertsit into an average gray scale intensity. The intensity of the yellowcolor developed on the paper test was consistent and proportionalto the amount of artesunate present in the sample. With artesu-nate concentrations ranging from 0.0 to 20 mg/mL, a linearcalibration plot was obtained with a detection limit of 0.98 mg/mL.

2. Experimental

2.1. Paper microchip fabrication

The microchip was fabricated using Whatman filter paper(Grade 3 1003-055) or glass microfiber GF/B (both from SigmaAldrich, St. Louis, MO 631778, USA). The filter paper is cut intocircular test pads using a hole-punch (8-mm diameter). Wedesigned the paper microchip to enable fluid transport verticallyby stacking the paper test pads. Each of the paper cut out has beenpretreated by spotting them with different reagents and allowedto dry. We assembled the pads as shown in Fig. 1, and securedthe stack on a plastic holder made of polymethyl methacrylate(PMMA). Four holes of 8-mm diameter were drilled into thePMMA piece, and the pad-carrier was laminated with a plasticcover on one side to secure the test pads in place. The PMMA piecewas then placed on the Artesunate Colorimetric Field Test couponprovided with a color chart guide (Fig. 1).

2.2. Reagent preparation

Sodium chloride and sodium citrate were purchased fromMallinckrodt (Phillipsburg, NJ, USA). Fast red TR salt, artesunate,artemisinin, NaOH, and citric acid were purchased from Sigma(St Louis, MO, USA). The reagents were prepared and loaded on thepad stack. We spotted 3 M citric acid buffer (25 μL) and 5 mg/mLfast red TR solution (25 μL) on the 8-mm paper pad stack. Thespotted reagents were then allowed to air dry at room temperature.

2.3. Antimalarial drug testing

A series of artesunate solutions ranging from 1 to 20 mg/mLwere prepared by diluting a 60 mg/mL stock solution in 1 M NaOH.We selected three commercially available drugs, one contain-ing artesunate (Guilin Pharmaceutical, China), and two otheranti-malarial drugs containing artemeter (Artem, KunmingPharmaceutical Corp, China) and dihydroartemisinin (Arterakine,Pharbaco Central Pharmaceutical, Vietnam). The drugs were pre-pared at 5 mg/mL and tested against a standard.

2.4. Colorimetric assay with iPhone

Color measurement of each chip was obtained by using aniPhone app (ColorAssist). Sample was applied to the paper chipand allowed to incubate for 5 min for color development. Analysis

of the image was performed by measuring the RGB value of eachspot, and converted to average gray intensity. A plot of averagegray intensity versus substrate concentration was generatedusing Excel.

3. Results and discussion

Artesunate is the most widely used artemisinin derivativethroughout the world for the treatment of severe malaria infec-tions and multi-drug resistant infections caused by P. falciparum,yet more than one-third of the antimalarial drugs tested over thepast decade in Southeast Asia and sub-Saharan Africa were eitheradulterated or of poor quality [19–22], containing the incorrectchemical formulation. Therefore, identification of authentic arte-sunate is essential to prevent needless fatalities. A colorimetric testhas been developed by the CDC to determine the authenticity ofartesunate in tablets. The test is based on the reaction of an alkalidecomposition product of artesunate with a diazonium salt, fastred TR (FRTR), that results in an appearance of a yellow color in thepresence of artesunate. The colored reaction product can bemeasured by spectrophotometry at 420 nm. The test is specificto artesunate when conducted at pH 4 [4].

In this work, we used Whatman filter paper to fabricatean effective, portable, low-cost paper-based test kit. The testkit consisted of four layers of paper, each containing differentreagents, similar to what is used in lateral flow chip design. Each ofthe reagents was spotted on the paper and air-dried at roomtemperature. Since artesunate specificity is pH dependent, it wasnecessary to maintain the reaction at the proper buffer conditionin each layer: from sample introduction to the final indicator layer.The test chip was assembled by stacking the paper circles toenable fluid transport vertically. We found that the vertical designyielded optimal pH control, and maintained pH 4 as the samplesolution traveled down along the paper layers. Here, the sampleflowed via capillary action through test zones pretreated with theappropriate reagents along the test path. Flow systems such asshown here are ideal platforms for sequential chemical processing.In addition, the vertical flow design resulted in reduced samplesize and a more consistent yellow color product, in contrast toa more uneven band spreading found in the lateral flow design.

The specificity of FRTR colorimetric detection is dependenton the pH of the reaction. Among a list of other anti-malarialsincluding artemisinin, artemether, chloroquine, quinine, prima-quine, sulfadoxine and pyrimethamine, artesunate is the only drugthat gives a yellow color at a pH of 4 [23]. Therefore, the reactionmust be maintained at pH 4 in order to be specific to artesunate,and not give positive test results in the presence of certain otheranti-malarials. We tested the specificity of the artesunate test onanother anti-malarial drug, artemisinin. When the reaction wasperformed at pH 6.5, both the artesunate and artemisinin samplesturned orange, indicating that the pH of the test was incorrect,resulting in invalid measurements. At pH 4, the artemisininyielded no color product, while the artesunate generated a yellowcolor product. This test validated the specificity of the artesunatetest at pH 4, as shown in Fig. 2.

The same reaction scheme was used to develop the papermicrochip. In order to maintain the solution at pH 4 prior tomixing with FRTR, the volume ratio of the 1 M NaOH:3 M citricacid buffer was adjusted. For consistent test conditions, we usedequal size of paper circles of 8-mm diameter, and spotted themwith the citric acid and FRTR reagents. Each circle received equalvolumes of the reagents (25 μL) and was allowed to dry at roomtemperature. When fully dried, the paper circles were assembledas shown in Fig. 1, and were ready for use. Next, 100 μL of the

Fig. 2. pH dependence of FRTR colorimetric test in solution. Artemisinin turns anorange color at pH 6.5, while giving no color at a pH of 4. Artesunate turns anorange color at pH 6.5, while giving a yellow color at a pH of 4.

M.T. Koesdjojo et al. / Talanta 130 (2014) 122–127124

dissolved artesunate tablet was pipetted on the test device andallowed to react for 5 min for the color to fully develop.

The chip design was evaluated for its specificity to artesunate.Both artesunate and artemisinin were applied to paper. The far leftimage in Fig. 3 shows a test result for artemisinin as added toa single paper layer containing both the dried citric acid buffer andthe FRTR reagent (equal volumes). An orange color developed,indicating that the test was not done at the proper pH of 4. Thenext image shows a test where artemisinin was added to a four-layered chip. The first layer was a blank serving as an absorbentpad, and was followed by two layers of dried citric acid, with abottom layer containing dried FRTR reagent. No color developed asexpected when the pH of the reaction occurs at the desired pH of4. The four-layered chip allowed for the basic sample solution toreact with the citric acid layers, and adjusted it to the appropriatepH of 4 as the analyte reached and the bottom layer containing theFRTR reagent. An orange color developed when artesunate wasadded to the one-layer chip (containing dried citric acid buffer anddried FRTR), indicating a pH above 4 when the reaction occurred. Ayellow color developed as expected when artesunate was added tothe four-layer chip at the desired pH of 4 as shown in the lastpicture. An additional pH test was also conducted by adding the1 M NaOH solution onto a two-layered paper chip that waspreloaded with the citric acid buffer in each layer. The pH of thesolution was measured at the base of the paper chip, and wasconfirmed to give a pH reading between 3.5 and 4.0 (Fig. 3b).

We employed the fabricated paper chip for the detection andmeasurement of artesunate solutions of varying concentration. Forthis test, the four-layer paper device design, as shown in Fig. 1, wasused. The first layer contained no reagent and served as anabsorbent pad for addition of sample solution. The second andthird layers contained the dry citric acid buffer reagent to adjustthe pH of the basic sample solution (artesunate dissolved in 1 MNaOH) to the proper pH of 4. The final layer contained the dryFRTR reagent required for the colorimetric assay. The test showedthat it was capable of detecting the presence (or absence) ofartesunate in the sample as indicated by the color change. Aftera 5 min incubation time, the yellow color was developed. Thedetermination of artesunate concentration was established using acalibrated color chart guide. The test showed no color develop-ment when blank was added on the paper chip indicating theabsence of artesunate in the sample. A yellow color was producedwhen artesunate was present, with darker colors indicating higherartesunate concentration in the tablet. The increasing color inten-sity developed on the paper chip was consistent with theincreased amount of artesunate in solution with 0.5 mg/mL beingthe lowest and 10 mg/mL the highest concentration tested. As canbe seen from Fig. 4, the more concentrated artesunate solutionresulted in a darker yellow color, with decreasing intensity as theconcentration of artesunate is decreased. We were able to obtainreproducible color products for detection and measurement ofartesunate using the test paper.

The performance of the paper test kit was assessed usingcommercially obtained anti-malarial drugs that are currentlybeing used to treat patients. At the time of this study, we had noaccess to real counterfeit drugs to evaluate the paper test kit.Therefore, as proof of concept, the test was conducted on other

antimalarial drugs to demonstrate the specificity of the test kit inidentifying artesunate and its accuracy in measuring artesunateconcentration. We selected three commercially available drugs,one containing artesunate (Guilin Pharmaceutical), and two otheranti-malarial drugs containing artemeter (Artem) and dihydroar-temisinin (Arterakine). From the three drugs evaluated, only thesample containing artesunate from Guilin Pharmaceutical testedpositive for artesunate while the other two drugs showed theabsence of artesunate in both samples. All three antimalarial drugswere prepared at 5 mg/mL for testing. Using the paper test, wewere able to determine the presence of artesunate in the Guilindrug at the correct concentration range based on the color chartreading (Fig. 5). The test was capable of successfully detecting thepresence (or absence) of artesunate in real samples and theirrelative concentration based on a calibrated color chart.

The color chart assay proved to be a reliable method fordetecting artesunate, especially when used as a field-test kit inremote areas. For a more accurate analysis, we utilized a coloranalyzer on a camera phone that measured the gray scale intensityof the yellow color developed on the test paper. After the reactionwas complete, a digital photo was taken of the paper chip. Theimage was analyzed by measuring the intensity of yellow colordeveloped on the paper circle. We used an iPhone cameraequipped with the “ColorAssist” application to obtain red, green,and blue component data, and then convert to gray scale intensity.ColorAssist is an app that uses the iPhone or iPad camera tocapture RGB values in real time. The iPhone camera was positionedat a set distance from the test paper to obtain reproducible andconsistent pictures throughout the analysis. A plot of the artesu-nate concentration versus the average gray scale intensity wasgenerated as shown in Fig. 6. The data showed that the intensity ofthe yellow color developed on the paper test coupon was con-sistent and proportional to the amount of artesunate present inthe sample. With artesunate concentrations ranging from 0.0 to20 mg/mL, a linear calibration plot was obtained with a detectionlimit of 0.98 mg/mL (n¼6).

4. Conclusions

We have demonstrated proof-of-concept for detecting counter-feit drugs using a paper based colorimetric assay. This approach issuitable and can serve as a rapid, simple, and cost effective meansthat can be used by the end user to test the authenticity of the drugin question. In addition, the simple detection technique couldbenefit government officials and local pharmacies as a rapidscreening method to combat the growing problem of drug counter-feiting today. Our future studies will include optimizing the assay toachieve lower limit of detection, and improved image analysis by

Fig. 3. (a) pH dependence of FRTR colorimetric test on paper. (b) pH test of 1 M NaOH solution added on two layered paper chips that were preloaded with citric acid buffer(left) and a blank (right).

Fig. 4. Semi-quantitative data utilizing the designed paper test device. From left toright: decreasing concentrations of artesunate from 10 mg/mL to 0 mg/mL.

M.T. Koesdjojo et al. / Talanta 130 (2014) 122–127 125

developing a paper device and a color sensor that further enhancesensitivity of the measurement.

Artesunate has been the most widely used artemisinin derivativethroughout the world for the treatment of severe malaria infectionsand multi-drug resistant infections caused by P. falciparum. Morethan one-third of the antimalarial drugs tested over the past decadein Southeast Asia and sub-Saharan Africa have been shown to becounterfeit or of poor quality, containing the incorrect chemicalformulation. This could undo decades of international effort to fightmalaria, and will promote the spread of resistance againstartemisin-based drugs. Therefore, identification of authentic arte-sunate is essential to prevent fatalities. We developed a paper-basedcolorimetric assay for the detection of antimalarial counterfeits. Thetest is capable of successfully detecting the presence (or absence) ofartesunate in real samples and their relative concentration based ona calibrated color chart. This paper based assay, together with acalibrated color chart guide would be an extremely useful field testkit in providing a semi-quantitative analysis of the artesunate drugin question. Using a color analyzer on an iPhone, we were able tosuccessfully perform a more quantitative colorimetric reading. Adigital picture was taken and analyzed based on its gray intensitythat linearly correlated proportional to its artesunate concentration.Future development of the device will include lower limit ofdetection of the assay and integration of a calibrated color analyzerthat would enable a direct quantitative measurement by scanningthe paper device.

Paper-based microfluidics assays have tremendous potential inthe field of medical- and drug-diagnostics. We envision that thistechnology will open up a new approach and platform fordetecting counterfeit drugs. Our future work will focus on improv-ing the architecture and expanding the range of applications of thedevice proposed above to achieve faster and more accurate

determination of counterfeit pharmaceuticals. In addition, a sim-ple color reader application (app) that will allow users to deter-mine the concentration values of the pictures taken is currentlybeing developed in our lab. Because this is a platform develop-ment effort, once established, the technology can be utilized toaddress a wide range of applications. This would further demon-strate the advantage of this platform over current lab-basedmethods of drug detection, which require extensive sample pre-paration, manual processing steps and expensive instrumentation.

Acknowledgment

The authors acknowledge Neal Sleszynski for the insight andhelpful discussion on the fabrication and detection techniquesused in this study. We gratefully appreciate Sumate Pengpumkiatfor his editorial assistance.

References

[1] World Health Organization, artesunate for the treatment of severe malaria.[2] P. Newton, S. Proux, M. Green, Lancet 357 (2001) 1948–1950.[3] P.N. Newton, R. McGready, F. Fernandez, M.D. Green, M. Sunjio, et al., PloS

Med. 3 (2006) 752–755.[4] A. Marshall, Smithsonian (2009), /http://www.smithsonianmag.com/people-

places/the-fatal-consequences-of-counterfeit-drugs-139422027S.[5] M.D. Green, D.L. Mount, R.A. Wirtz, N.J. White, J. Pharm. Biomed. Anal. 24

(2000) 65–70.[6] P.N. Newton, M.D. Green, F.M. Fernandez, N.P.J Day, N.J. White, Lancet Infect.

Dis. 6 (2006) 602–613.[7] K.A. Hall, P.N. Newton, M.D. Green, M.D. Veij, P. Vandanabeele, D. Pizzanelli,

M. Mayxay, A. Dondorp, F. Fernandez, Am. J. Trop. Med. Hyg. 75 (2006)804–811.

[8] N. Lindegardh, A.M. Dondorp, P. Singhasivanon, N.J. White, N.P.J. Day, J. Pharm.Biomed. Anal. 45 (2007) 149–153.

Fig. 5. (a) Semi-quantitative data utilizing the designed paper test device. From left to right is increasing concentrations of artesunate from 0 mg/mL to 10 mg/mL. The realanti-malarial drugs were prepared at 5 mg/mL, tested and compared with the standard. (b) The Artesunate Field Test Kit.

Fig. 6. Plot for the mean intensity of the color produced versus the artesunate concentration with iPhone ColorAssist app. The plot shows very good linearity(R2 value¼0.9945) with detection limit of 0.98 mg/mL.

M.T. Koesdjojo et al. / Talanta 130 (2014) 122–127126

[9] A.A. Nuhu, J. Appl. Pharm. Sci. 5 (2011) 6–13.[10] M. Green, D. Mount, G. Todd, Analyst 120 (1995) 2623–2626.[11] D. Mount, B. Nahlen, L. Patchen, F. Churchill, Bull. World Health Organ. (1989)

295–300.[12] A.W. Martinez, S.T. Phillips, B.J. Wiley, M. Gupta, G.M. Whitesides, Lab Chip 8

(2008) 2146–2150.[13] Y. Lu, W. Shi, L. Jiang, J. Qin, B. Lin, Electrophoresis 30 (2009) 1497–1500.[14] E. Carrilho, A.W. Martinez, G.M. Whitesides, Anal. Chem. 81 (2009)

7091–7095.[15] X. Li, J. Tian, T. Nguyen, W. Shen, Anal. Chem. 80 (2008) 9131–9134.

[16] K. Abe, K. Suzuki, D. Citterio, Anal. Chem. 80 (2008) 6928–6934.[17] E.M. Fenton, M.R. Mascareñas, G.P. Lopez, S.S. Sibbett, ACS Appl. Mater.

Interfaces 1 (2009) 124–129.[18] P. Newton, Wellcome Trust 2001: fake artesunate in Southeast Asia (2001).[19] L.K. Basco, Am. J. Trop. Med. Hyg. 70 (2004) 245–250.[20] B. Daviss, Scientist 19 (2005) 42–43.[21] N.J. White, J. Clin. Invest. 113 (2004) 1084–1092.[22] P. Aldhous, Nature 434 (2005) 132–136.[23] M.D. Green, D.L. Mount, R.A. Wirtz, Trop. Med. Int. Health 6 (2001) 980–982.

M.T. Koesdjojo et al. / Talanta 130 (2014) 122–127 127